Lecture 6: Model Answer
Nettle et al (2012) tested if signs places by cycle racks reduced bike thefts. Their theory was based on previous research which showed that signage including eyes encouraged social norm following.
Signs were placed above cycles racks in three campus locations. These signs featured a large picture of a pair of male eyes, the slogan "Cycle Thieves we are watching you", the University security services logo and the words "operation crackdown". Cycle thefts were monitored for the year before and after sign placement, in these locations, and in control locations also around campus. The independent variable was sign presence. The dependent variable was thefts number. This was a controlled trial of an intervention.
Overall there was a similar number of thefts on campus before and after the placement of the signs (70 vs 68). In the experimental locations, with signs, thefts reduced from 39 to 15. In the control locations, without signs, thefts increased from 31 to 51. There was a highly significant interaction between presence of the sign and time. The introduction of the signs reduced thefts at those locations, increasing them at the other locations.
An odds ratio (O.R.) was calculated which compared the ratio of change in the control location to the ratio of change in the experimental location. The O.R. was 4.28 (95% confidence interval 2.04–8.98).
The O.R. was also significantly above 1 for both halves of the year after sign placement.
The O.R. incorporates the displacement of thefts from the experimental locations to the control locations. An intervention which does not reduce crime may have a highly significant odds ratio.
CONCLUSION, IMPLICATIONS AND CRITICISMS
The authors suggest that thefts were displaced from locations with signs to locations without signs.
This study is small, but highly ecologically valid. Although only three locations, and all on one campus, were used to site signs, the use of real crime statistics means that the dependent variable of the study has direct relevance to an important issue.
A number of a questions remain:
1. What about the signs caused the effect?
The signs contained a number of elements - warnings, slogans, logos and eyes. Because only one type of sign was used it is not possible to infer which elements caused the effect. Although the authors acknowledge this, it could still be said that focussing their review on research about the effect of eyes is misleading - this study does not directly support the power of pictures of eyes alone to change behaviour.
2. Can the effect be used to reduce overall crime?
The odds ratio statistic disguises that the signs appeared to displace crime rather than reduce it. It remains to be seen if widespread deployment of the signs would actually reduce overall crime levels.
3. Would the effect last?
It is highly plausible that criminals would adjust to the presence of signs, especially if they learnt that the signs were false warnings of actual surveillance. The authors report no attenuation of the effect over time, but this may have been due to low statistical power. Furthermore, the authors discuss the issue of the number of individuals affected by the signs - this interacts with whether the longevity of the effect can be assessed. If there are only a few thieves stealing bikes it would be easier to find evidence of an attenuation - if it existed. If there were many thieves stealing bikes, you would not expect to see adjustment in behaviour so easily.
4. What was the mechanism of the effect?
It is difficult to know what about the signs caused the effect. Nor is it clear which psychological mechanisms operated to bring about an effect. The authors cite earlier work showing that signs could influence conformity with social norms. Social norm conformity may not be a large factor in the decision of someone to steal a bike. More plausibly, bike thieves worry about being caught, which the signs seems well designed to influence. The authors draw on a research claiming non-conscious influence of social cues, but this study leaves open the possibility that thieves were affected by an entirely deliberative consideration. It is not possible to make any inference about the influence of subtle social cues from this study.
Another possibility, which was not considered by the authors, is that the signs influence cyclists rather than cycle thieves. It is possible that security-conscious cyclists chose differentially to lock their bikes in locations with signs. This would leave poorly locked bikes in the non-signed locations, and thus raise the level of thefts from these locations.
Overall, this study is highly suggestive, and is commendable to using a realistic intervention in a every day environment. However, further tests are needed to assess the efficacious elements of the sign, the psychological mechanisms of effect and - following from this - the possibilities for lasting reduction in overall crime levels.
Suggested future research could address these four areas of criticism. For example:
Varying the sign elements could test which elements drive the effect.
Adding context of national crime statistics might inform whether this intervention could reduce crime. If crime overall was increasing, the consistent number of thefts in the 'after' part of the study would actually be a relative reduction.
The crime data could be collected for many subsequent years, to see if the effect lasted. Or, alternatively, the same study could be carried out in locations with a low number of cycle thieves (e.g. remote communities?) - it would be easiest to observe any evidence of adaptation in these circumstances.
Finally, the mechanism of action could be tested by locking bikes in identical ways in signed and non-signed locations and measuring their likelihood of being stolen. This would control for the effect occurring via the cyclists' actions. If the mechanism is deliberative reasoning by the thieves, then signs which are equally eye catching and contain a warning would have the same effect as the signs with the eyes.
Nettle, D., Nott, K., & Bateson, M. (2012) “Cycle Thieves, We Are Watching You”: Impact of a Simple Signage Intervention against Bicycle Theft. PloS one, 7(12), e51738.